This could indeed be it. The baseline for Bullseye still is a basic 686. Additionally, the lowest Linux kernel for i386 is linux-image-686 which is configured for Geode. Still, the bug just appeared a few days ago when the package in Testing was updated. This seems like a regression.
$ lscpu Architecture: i586 CPU op-mode(s): 32-bit Byte Order: Little Endian Address sizes: 32 bits physical, 32 bits virtual CPU(s): 1 On-line CPU(s) list: 0 Thread(s) per core: 1 Core(s) per socket: 1 Socket(s): 1 Vendor ID: AuthenticAMD CPU family: 5 Model: 10 Model name: Geode(TM) Integrated Processor by AMD PCS Stepping: 2 CPU MHz: 498.044 BogoMIPS: 996.08 L1d cache: 57 KiB L1i cache: 57 KiB L2 cache: 128 KiB Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Spec store bypass: Vulnerable Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Full generic retpoline, STIBP disabled, RSB filling Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected Flags: fpu de pse tsc msr cx8 sep pge cmov clflush mmx mmxext 3dnowext 3dnow cpuid 3dnowprefetch vmmcall ke 10. maalisk. 2021 klo 12.49 Julien Cristau (jcris...@debian.org) kirjoitti: > > This looks relevant: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84148 > > Cheers, > Julien > > On Wed, Mar 10, 2021 at 11:20:25AM +0100, Bernhard Übelacker wrote: > > Hello Martin-Éric, > > without being involved in packaging fwupd I tried to > > have a look at this issue. > > > > I could not reproduce it inside a i386 qemu VM (not even > > with "-cpu pentium"). Have not tested on real hardware. > > > > > > Looking up the endbr32 instruction, it seems it belongs to something > > called "Control-flow Enforcement Technology" (CET, indirect branch) [1]. > > > > The opcode for this instruction got selected to run on old > > CPUs as NOP, but it looks like your CPU handles it differently. > > From the system name it is some "geode" CPU? > > > > (In [2] someone mentions also a illegal instruction > > for a geode CPU with the endbr32 instruction.) > > > > > > Maybe you could add to this bug report the output of > > 'lscpu' or 'cat /proc/cpuinfo' ? > > > > Then the maintainer might be able to tell if this CPU > > meets the Debian baseline requirements for bullseye. > > ([3], maybe outdated? Is there a better "baseline" description?) > > > > > > Kind regards, > > Bernhard > > > > [1] > > https://www.linuxplumbersconf.org/event/2/contributions/147/attachments/72/83/CET-LPC-2018.pdf > > [2] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84148#c3 > > [3] https://wiki.debian.org/ArchitectureSpecificsMemo#i386-1 > >
